INDEX

Explanations

formatting specifiers typically seen in programming and output functions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Халык

-0.65

 Biôgrafia

-0.50

ViewFeatures

-0.47

InitStruct

-0.45

 plaisir

-0.45

 препратки

-0.44

cedure

-0.43



-0.42

AGUE

-0.42

motivation

-0.41

POSITIVE LOGITS

lld

0.52

llu

0.48

zu

0.47

ld

0.46

lu

0.46

0.40

 Infórmanos

0.39

Lf

0.38

zd

0.37

lf

0.37

Activations Density 0.194%

formatting specifiers typically seen in programming and output functions

No Comments

No Known Activations